基于神经网络的电动转向系统助力特性研究  

A Study on Assist Characteristic of Electric Power Steering System Based on Neural Network

作　　者：尚喆[1] 许镇琳[1] 王豪[1] 张海华[1] 

机构地区：[1]天津大学电气与自动化工程学院,天津300072

出　　处：《汽车工程》2004年第3期319-321,共3页Automotive Engineering

摘　　要：基于综合目标函数的误差反向传播学习算法 ,设计了用于该离散化转向助力特性的BP神经网络结构 ,进行了离线训练 ,并在电动转向综合试验台上实现了在线控制。实现了全车速范围的非线性转向助力 ,克服了转向助力盲区。Based on error back propagation algorithm of overall objective function,a structure of BP neural network used for discrete steering assisting is designed,the off line training is performed,and the on line control is implemented on EPS test rig.So the nonlinear steering assist in full speed range is realized and the blind zone is eliminated.

关 键 词：神经网络 电动转向系统 助力特性 汽车 误差反向传播学习算法 

分 类 号：U463.6[机械工程—车辆工程]